检索结果-内蒙古大学图书馆

Visual and structural feature combination in an interactive machine learning system for medical image segmentation

MACHINE LEARNING WITH APPLICATIONS 2022年 8卷

作者： Galisot, Gaetan Ramel, Jean -Yves Brouard, Thierry Chaillou, Elodie Serres, Barthelemy Univ Tours LIFAT 64 Ave Jean Portalis F-37200 Tours France Univ Tours CNRS IFCE INRAE F-37380 Nouzilly France Univ Tours LIFAT ILIAD3 64 Ave Jean Portalis F-37200 Tours France

Currently, Convolutional Neural Networks achieve good performance in automatic image segmentation situations;however, they have not demonstrated sufficiently accurate and robust results in the case of more general and interactive systems. Also, they have been designed specifically for visual features and cannot integrate enough anatomical knowledge inside the learned models they produce. To address these problems, we propose a novel machine -learning -based framework for interactive medical image segmentation. The proposed method incorporates local anatomical knowledge learning capabilities into a bounding box -based segmentation pipeline. Region specific voxel classifiers can be learned and combined to make the model adaptive to different anatomical structures or image modalities. In addition, a spatial relationship learning mechanism is integrated to capture and use additional topological (anatomical) information. New learning procedures have been defined to integrate both types of information (visual features to characterize each substructure and spatial relationships for a relative positioning between the substructures) in a unified model. during incremental and interactive segmentation, local substructures are localized one by one, enabling partial image segmentation. Bounding box positioning within the entire image is performed automatically using previously learned spatial relationships or by the user when necessary. Inside each bounding box, atlas -based methods or CNNs that are dedicated to each substructure can be applied to automatically obtain each local segmentation. Experimental results show that (1) the proposed model is robust for segmenting objects with a small amount of training images;(2) the accuracy is similar to other methods but allows partial segmentation without requiring a global registration;and (3) the proposed method leads to accurate results with fewer user interactions and less user time than traditional interactive segmentation method

关键词： 3d image segmentation Machine learning Interactive method Spatial relationship Atlas Brain images

来源：评论

学校读者我要写书评

暂无评论

Automatic three-dimensional reconstruction of subsurface defects by segmenting ultrasonic point cloud

引用

JOURNAL OF THE TAIWAN INSTITUTE OF CHEMICAL ENGINEERS 2021年 120卷 24-32页

作者： Zheng, Kaiyi Yao, Yuan Jiangsu Univ Sch Food & Biol Engn Key Lab Modern Agr Equipment & Technol Zhenjiang 212013 Jiangsu Peoples R China Natl Tsing Hua Univ Dept Chem Engn Hsinchu 30013 Taiwan

Ultrasonic inspection is a widely accepted nondestructive method for subsurface defect detection of materials. In the past years, a number of signal processing techniques have been adopted to improve defect visibility. However, an automatic identification method that reconstructs the three-dimensional (3d) shapes of the defects from the ultrasonic data is still desired. In this work, a 3d minimum spanning tree (3d-MST) method is proposed, which treats the ultrasonic data as a point cloud and segment it by finding a tree whose sum of edge weights is as small as possible. The segmentation results give a clear indication of the positions, sizes, and shapes of the subsurface defects contained in the materials. It is also discussed that the segmentation results can be improved by adopting robust normalization as a pretreatment step before conducting 3d-MST. The feasibility of the proposed method is illustrated by case studies on a polymer composite and a food product. (c) 2021 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

关键词： 3d image segmentation Cluster analysis Minimum spanning tree (MST) Nondestructive testing Ultrasonic inspection

来源：评论

学校读者我要写书评

暂无评论

Topological active volumes: A topology-adaptive deformable model for volume segmentation

引用

PATTERN RECOGNITION 2010年第1期43卷 255-266页

作者： Barreira, N. Penedo, M. G. Cohen, L. Ortega, M. Univ A Coruna Dept Comp Sci La Coruna Spain Univ Paris 09 CEREMADE CNRS Appl Math & Image Anal Grp F-75775 Paris 16 France

This paper proposes a generic methodology for segmentation and reconstruction of volumetric datasets based on a deformable model, the topological active volumes (TAV). This model, based on a polyhedral mesh, integrates features of region-based and boundary-based segmentation methods in order to fit the contours of the objects and model its inner topology. Moreover, it implements automatic procedures, the so-called topological changes, that alter the mesh structure and allow the segmentation of complex features such as pronounced curvatures or holes, as well as the detection of several objects in the scene. This work presents the TAV model and the segmentation methodology and explains how the changes in the TAV structure can improve the adjustment process. In particular, it is focused on the increase of the mesh density in complex image areas in order to improve the adjustment to object surfaces. The Suitability of the mesh structure and the segmentation methodology is analyzed and the accuracy of the proposed model is proved with both synthetic and real images. (C) 2009 Elsevier Ltd. All rights reserved.

关键词： 3d image segmentation Topological active volumes Adaptive topology

来源：评论

学校读者我要写书评

暂无评论

Literature review: efficient deep neural networks techniques for medical image analysis

引用

NEURAL COMPUTING & APPLICATIONS 2022年第8期34卷 5791-5812页

作者： Abdou, Mohamed A. City Sci Res & Technol Applicat Informat Res Inst Knowledge Based Syst Dept Alexandria Egypt

Significant evolution in deep learning took place in 2010, when software developers started using graphical processing units for general-purpose applications. From that date, the deep neural network (dNN) started progressive steps across different applications ranging from natural language processing to hyperspectral image processing. The convolutional neural network (CNN) mostly triggers the interest, as it is considered one of the most powerful ways to learn useful representations of images and other structured data. The revolution of dNNs in medical imaging (MI) came in 2012, when Li launched imageNet, a free database of more than 14 million labeled medical images. This state-of-the-art work presents a comprehensive study for the recent dNNs research directions applied in MI analysis. Clinical and pathological analysis through a selected patch of most cited researches is introduced. It will be shown how dNNs are able to tackle medical problems: classification, detection, localization, segmentation, and automatic diagnosis. datasets comprises a range of imaging technologies: X-Ray, MRI, CT, Ultrasound, PET, Fluorescene Angiography, and even photographic images. This work surveys different patterns of dNNs and focuses somehow on the CNN, which offers an outstanding percentage of solutions compared to other dNNs structures. CNN emphasizes image features and has well-known architectures. On the other hand, limitations beyond dNNs training and execution time will be explained. Problems related to data augmentation and image annotation will be analyzed among a multiple of high standard publications. Finally, a comparative study of existing software frameworks supporting dNNs and future research directions in the area will be presented. From all presented works it could be deduced that the use of dNNs in healthcare is still in its early stages, there are strong initiatives in academia and industry to pursue healthcare projects based on dNNs.

关键词： Convolutional neural networks Restricted Boltzmann machine Pathological image analysis image localization Clinical image diagnosis 3d image segmentation

来源：评论

学校读者我要写书评

暂无评论

SEMIAUTOMATIC 3-dIMENSIONAL segmentation METHOd dISARTICULATION OF BONE STRUCTURES ON SPIRAL COMPUTEd-TOMOGRAPHY imageS

引用

JOURNAL OF dIGITAL IMAGING 1995年第4期8卷 156-161页

作者： VANCLEYNENBREUGEL, J KRATKA, d BERBEN, L SMET, MH MARCHAL, G SUETENS, P UNIV HOSP GASTHUISBERG DEPT RADIOLB-3000 LOUVAINBELGIUM

The use of binary thresholding for segmenting bone structures on spiral computed tomography images is negatively influenced by partial volume effects (PVEs) induced by the image acquisition. PVE leads to mixed voxels, making the binary decision ''bone'' or ''non-bone'' a difficult one to take. As a result, two distinct bone structures that are close to each other will often appear to be connected by this method. A typical example consists of ''acetabulum/femural head'' pairs in the pelvic region. To separate them, a clinical user must interactively draw a disarticulation line. This procedure is time consuming (often interaction in 50 slices is needed) and leads to unsmooth visualization of the disarticulated areas (by three-dimensional [3d] rendering techniques). We developed a semiautomatic cutting algorithm that leads to smooth disarticulated surfaces and considerably decreases the amount of user interaction. A sheet detection operator is applied to automatically separate bone structures. detected sheets are used as disarticulation lines. Postprocessing ensures that sheets not relevant for the application do not influence the resulting image. Our approach is encapsulated in an interactive segmentation environment based on thresholding and 30 connected-component labeling. Results are shown for pelvic region, wrist, and foot bone disarticulations. (C) 1995 by W.B. Saunders Company

关键词： 3d image segmentation 3d image USER INTERACTION COMPUTEd TOMOGRAPHY BONE segmentation

来源：评论

学校读者我要写书评

暂无评论

A variational model and its numerical solution for local, selective and automatic segmentation

引用

NUMERICAL ALGORITHMS 2014年第2期66卷 399-430页

作者： Rada, Lavdie Chen, Ke Univ Liverpool Ctr Math Imaging Tech Liverpool L69 7ZL Merseyside England Univ Liverpool Dept Math Sci Liverpool L69 7ZL Merseyside England

Variational region-based segmentation models can serve as effective tools for identifying all features and their boundaries in an image. To adapt such models to identify a local feature defined by geometric constraints, re-initializing iterations towards the feature offers a solution in some simple cases but does not in general lead to a reliable solution. This paper presents a dual level set model that is capable of automatically capturing a local feature of some interested region in three dimensions. An additive operator spitting method is developed for accelerating the solution process. Numerical tests show that the proposed model is robust in locally segmenting complex image structures.

关键词： image selective segmentation Level set functions Euler-Lagrange equation 3d image segmentation Operator spitting

来源：评论

学校读者我要写书评

暂无评论

CA-UNet segmentation Makes a Good Ischemic Stroke Risk Prediction

引用

INTERdISCIPLINARY SCIENCES-COMPUTATIONAL LIFE SCIENCES 2024年第1期16卷 58-72页

作者： Zhang, Yuqi Yu, Mengbo Tong, Chao Zhao, Yanqing Han, Jintao Beihang Univ Sch Comp Sci & Engn Beijing Peoples R China Beihang Univ State Key Lab Virtual Real Technol & Syst Beijing Peoples R China Peking Univ Third Hosp Dept Intervent Radiol & Vasc Surg Beijing Peoples R China

Stroke is still the World's second major factor of death, as well as the third major factor of death and disability. Ischemic stroke is a type of stroke, in which early detection and treatment are the keys to preventing ischemic strokes. However, due to the limitation of privacy protection and labeling difficulties, there are only a few studies on the intelligent automatic diagnosis of stroke or ischemic stroke, and the results are unsatisfactory. Therefore, we collect some data and propose a 3d carotid Computed Tomography Angiography (CTA) image segmentation model called CA-UNet for fully automated extraction of carotid arteries. We explore the number of down-sampling times applicable to carotid segmentation and design a multi-scale loss function to resolve the loss of detailed features during the process of down-sampling. Moreover, based on CA-Unet, we propose an ischemic stroke risk prediction model to predict the risk in patients using their 3d CTA images, electronic medical records, and medical history. We have validated the efficacy of our segmentation model and prediction model through comparison tests. Our method can provide reliable diagnoses and results that benefit patients and medical professionals.

关键词： Ischemic stroke Carotid Stroke risk prediction deep learning 3d image segmentation

来源：评论

学校读者我要写书评

暂无评论

Statistical shape model building method using surface registration and model prototype

引用

OPTICS ANd LASER TECHNOLOGY 2019年 110卷 234-238页

作者： Li, Guangxu Wu, Jiaqi Xiao, Zhitao Kim, Hyoung Seop Ogunbona, Philip O. Tianjin Polytech Univ Sch Informat & Commun Engn Tianjin Peoples R China Kyushu Inst Technol Dept Control Engn Fukuoka Fukuoka Japan Univ Wollongong Sch Comp & Informat Technol Wollongong NSW Australia

Automatic segmentation of organs from medical images is indispensable for the computer-assisted medical applications. Statistical Shape Models (SSMs) based scheme has been developed as an accurate and robust approach for extraction of anatomical structures, in which a crucial step is the need to place the sampled points (landmarks) with well corresponding across the whole training set. On the one hand, the correspondence of landmarks is related the quality of shape model. On the other hand, in clinical application some key positions of landmarks should be specified by physicians referring to the anatomic structure. In this paper, we develop an interactive method to build SSM that the landmark distribution can be modified manually without influencing the model quality. We extend an existing remeshing method to produce a model prototype in advance and surface features driven registration to insure the universal optimization of correspondence. The key landmarks are fixed during the prototype generation. We experimented and evaluated the proposed SSM method for lung regions, the deformations of which are considerable large. (C) 2017 Elsevier Ltd. All rights reserved.

关键词： Statistical shape model Landmarks correspondence Mesh registration 3d image segmentation

来源：评论

学校读者我要写书评

暂无评论

Intelligent environment for advanced brain imaging: multi-agent system for an automated Alzheimer diagnosis

引用

EVOLUTIONARY INTELLIGENCE 2021年第4期14卷 1523-1538页

作者： Allioui, Hanane Sadgal, Mohamed Elfazziki, Aziz Cadi Ayyad Univ Fac Sci Semlalia Comp Sci Dept BP 2390 Marrakech Morocco

Over decades Alzheimer's disease (Ad) researches presented increasing challenges. However, various methods were proposed to detect Ad including image processing. This paper presents a concrete solution to diagnose Ad based on a multi-agent system (MAS). This approach highlights the importance of the cooperation paradigm within a robust system, in which all agents cooperate to accomplish the segmentation tasks. The exchanges between agents remain an essential part of the segmentation process. The original contribution of this paper is twofold: (1) To present an agent-based segmentation methodology by highlighting the main characteristics, advantages, and disadvantages of MAS. (2) To provide a usable solution by facilitating the detection of Ad while taking into account both the expertise and the requirements of specialists in the application domain. Ensuring a cooperative segmentation using the multi-agent system offers a strong point in terms of system stability as well as clarity of the process for physicians. For this, several tests have been carried out to prove the effectiveness of our work. The results ensure that the performance indices in our proposed method were higher.

关键词： Multi-agent system (MAS) 2d 3d image segmentation 2d 3d image processing Alzheimer disease (Ad) Healthcare

来源：评论

学校读者我要写书评

暂无评论

Axon tracing in the adult rat optic nerve and tract after intravitreal injection of MndPdP using a semiautomatic segmentation technique

引用

JOURNAL OF MAGNETIC RESONANCE IMAGING 2008年第1期27卷 34-42页

作者： Olsen, Oystein Thuen, Marte Berry, Martin Kovalev, Vassili Petrou, Maria Goa, Pal Erik Sandvig, Axel Haraldseth, Olav Brekken, Christian Sor Trondelag Univ Coll Dept Radiodiag N-7004 Trondheim Norway Norwegian Univ Sci & Technol Dept Circulat & Med Imaging N-7034 Trondheim Norway Univ Birmingham Queen Elizabeth Med Ctr Dept Med Birmingham W Midlands England Belarus Natl Acad Sci United Inst Informat Problems Minsk BELARUS Univ London Imperial Coll Sci Technol & Med Dept Elect & Elect Engn Commun & Signal Proc Grp London England St Olavs Hosp Dept Med Imaging Trondheim Norway Ullevaal Univ Hosp Expt Med Res Inst Lab Regenerat Neurobiol Oslo Norway

Purpose: To develop and validate an objective technique for 3d segmentation of manganese-enhanced MR images of the optic nerve/tract (ON) in adult rats to improve contrast-to-noise (CNR) calculations and use the technique to ascertain if manganese dipyridoxyl diphosphate (MndPdP) gives sufficient Mn2+ enhancement compared to MnCl2 when used for functional imaging of the visual pathway. Materials and Methods: Intravitreous injection of the manganese-releasing MR contrast agent MndPdP (30 nmol Mn2+) was performed to trace the ON in adult rats (n = 4). A positive control group of rats (n = 5) received a standard preparation of MnCl2 (200 nmol Mn2+), while gadodiamide (1500 nmol Gd3+) was administered in negative control rats (n = 2). An objective technique for 3d segmentation of the enhanced ON was developed. CNR profiles along the ON were calculated by resampling the 3d image-volume in 2d planes perpendicular to the Mn2+ enhanced ON in 0.2 mm steps, 4 min along the segmented ON measured from the lamina cribrosa. Results: The ON was successfully segmented and CNR calculated accurately within 2 minutes in a representative 3d MR image volume. Intravitreal MndPdP injection resulted in significant MRI contrast enhancement of the retina and ON after 12-24 hours similar to that of MnCl2 injection. Conclusion: 3d semiautomated image segmentation and the use of MndPdP can improve in vivo axon tracing based on MRI. Mn2+ was found to be released from MndPdP after intravitreal injection in sufficient amounts to obtain functional tracing of the adult rat primary visual pathway.

关键词： 3d image segmentation optic nerve magnetic resonance imaging MEMRI manganese rat

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：